Inundation discriminated using sun glint

Abstract

Inundation is linked to water, carbon, and energy budgets at landscape to global scales. We describe a new remote-sensing technique for identifying inundated areas based on the properties of the glitter-the strong, angular signature reflection that is characteristic of surface water and uncharacteristic of other cover types. We discriminated three cover types-vegetation emergent above inundated soils, open water, and noninundated cover types-from analysis of directional data collected in the red spectral band by the airborne POLDER (Polarization and Directionality of Earth's Reflectance) sensor. We found that values of the normalized difference vegetation index (NDVI) decreased dramatically in the glitter direction, providing an indication of surface water. Application of our new technique holds promise for mapping the seasonal and interannual extent of inundation, a key descriptor of wetlands hydrology.